This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. A gene expression event involves the activation/repression of a signal transduction cascade and the cis-regulatory elements responding to such a signal. Information about the gene expression system is usually collected in heterogeneous forms of high throughput experimental data, e.g. activity state of a signal transduction pathway is usually embodied as the concentration changes of the signaling molecules within the pathway;information of cis-regulatory elements is contained in gene promoter sequence data;and information of transcription events (resulted from interaction of the trans- and cis-regulatory components) is reflected in microarray data. In this project, we will develop statistical models to integrate information from the above data sources within a probabilistic graphical model framework, in which the components within the systems are represented as variables and their interactions are explicitly modeled as probabilistic relationships. Our overall hypothesis is that, through information integration, the proposed models will enhance our capability to decipher the mechanisms of the gene expression system. By applying the proposed statistical models on the composite data, we will test specific hypotheses such as: information integration enhances identification the groups of genes regulated by signal transduction pathways, and it facilitates elucidating the mechanisms by which sphingolipids regulate gene expression. The developed models will lay a foundation for future study of mechanisms of diseases such as diabetes and cancer.